A full-scale shift to AI is the destiny of JD.
In 2026, JD is undergoing a major shift in its strategic focus.
The broader context of this shift is as follows: from a financial performance perspective, JD recorded its first year-on-year quarterly revenue decline since its listing in the second quarter, while its profit saw a substantial rebound, with Non-GAAP net profit increasing by 21% year-on-year. Regarding this performance, Xu Ran, CEO of JD, stated that the company's profitability trajectory has reached a clear inflection point.
However, right as the profitability inflection point emerges, JD's strategic focus is also shifting: AI is playing an increasingly critical role in JD's business system.
In fact, since the second half of 2025, Richard Liu has personally endorsed JD's AI initiatives on multiple occasions: last September, he took the position of dean of the upgraded JD Exploration Research Institute, elevating AI to a "top leader's priority project"; at the 2026 APEC China Business Leaders Forum in June, he emphasized that "JD is not the strongest player in the industry in terms of large model R&D, but it boasts rich advantages in physical scenario data".
Apart from technology bets, there is also a series of arrangements that bear distinct personal imprints of Richard Liu.
For example, he announced that all of JD's AI technologies will be fully open to overseas partners, on the grounds that "technical barriers are essentially the exploitation of technology"; he launched the "Nirvana Plan", which aims to bring 700,000 couriers and warehouse workers back to classrooms in batches for training in robot maintenance and repair.
"In the future, all deliveries will be done by robots... but I will not let 700,000 of our colleagues lose their livelihoods and jobs," he emphasized. "For frontline employees replaced by machines, JD will not lay off a single one of them." The goal is to build the "largest operator in the physical world", and "20 years from now, JD will still be the enterprise with the largest number of employees in China".
This is a statement of responsibility from an entrepreneur — but at the same time, JD first needs to answer a series of practical questions: how to implement AI and support the positive cycle of JD's business growth.
Food delivery losses narrowed sharply, profit growth exceeded 20%
Judging from the overall financial performance of the second quarter, JD is entering a period of financial fluctuations brought by the strategic focus shift.
According to the 2026 Q2 and interim results released by JD Group on August 13, 2026, JD's revenue in the second quarter reached 346.4 billion yuan, down 2.9% year-on-year — marking the first negative quarterly revenue growth since JD's listing.
On the other side of the financial report, however, is a sharp reversal in profits.
Financial data shows that JD's operating profit in the second quarter was 4.5 billion yuan, compared with a loss of 900 million yuan in the same period last year; Non-GAAP net profit attributable to ordinary shareholders reached 8.9 billion yuan, up about 21% year-on-year.
From a business segmentation perspective, the core reason for the impact on JD's Q2 revenue is that revenue from electronics and home appliances, as its core business, stood at 157.9 billion yuan, down 11.8% year-on-year from 179 billion yuan in the same period last year.
It is worth clarifying that the decline in revenue of the electronics-related business in Q2 is against the backdrop of a high base from last year. In 2025, driven by the national subsidy policy for "trade-in of old products", JD's Q2 revenue last year achieved a 22.4% year-on-year increase. This year, as the subsidy effect faded, the elevated base from last year directly translated into this year's decline.
In addition, another factor contributing to the decline of this business is the impact of demand. From the consumer electronics industry perspective, in 2026, price hikes of upstream components such as storage and chips have been transmitted to the end market, pushing up the average transaction value of mobile phones and digital products, and the demand for device replacement has been genuinely suppressed.
In response, Xu Ran, CEO of JD, stated that affected by multiple factors including the high base effect from last year and the price increase of electronics categories caused by rising raw material costs, the electronics-related categories faced short-term pressure in the second quarter. But she also emphasized that with JD's strong supply chain capabilities and solid user mindshare in these categories, JD's electronics-related categories have continued to consolidate their market position, and the market share of various home appliance categories is steadily increasing.
Outside of this business, JD's other three business lines are all growing.
Specifically, revenue from general merchandise in Q2 reached 109.2 billion yuan, accounting for 31.5% of total revenue, up 5.6% year-on-year; revenue from platform and advertising services stood at 30.9 billion yuan, accounting for 8.9%, up 8.3% year-on-year; revenue from logistics and other services was 48.4 billion yuan, accounting for 14%, up 5.9% year-on-year.
Then, with revenue dropping by nearly 3% year-on-year, where did JD's over 20% profit growth in Q2 come from?
The answer is: reduced losses.
Judging from the profit performance of each business segment, in Q2, JD Retail recorded an operating profit of about 13.5 billion yuan, a slight year-on-year decline of about 3%, which matches the overall revenue performance of the company. At the same time, JD Logistics' operating profit reached about 2.26 billion yuan, up 15.6% year-on-year.
More importantly, the operating loss of the new business segment, which includes food delivery, Jingxi, JD Properties Development and overseas businesses, narrowed from 14.78 billion yuan in the same period last year to 9.85 billion yuan — meaning that almost all profit improvement came from the narrowing of losses in new businesses.
Food delivery contributed the majority of the improvement.
During the earnings call, Xu Ran, CEO of JD, disclosed that the food delivery business "reduced its total loss by more than 50% year-on-year, and per-order subsidies decreased significantly". Changes on the expense side echoed this: marketing expenses in Q2 dropped sharply by 24.8% year-on-year to 20.3 billion yuan, saving about 6.7 billion yuan in a single quarter, which the financial report attributed to "optimization of promotion investment for new businesses".
Therefore, from the overall business perspective, the fading of the capital-burning intensity in the food delivery war has become the single biggest variable driving the profit performance of JD's Q2 financial report.
Expand revenue scale and maintain profit growth
For JD, against the backdrop of its Q2 performance, promoting the expansion of the company's revenue scale and overall profit growth has become the top concern of the outside world.
For the electronics-related category business, Xu Ran, CEO of JD, emphasized during the earnings call that starting from the third quarter, the impact of the high base from the national subsidy policy will be gradually digested; as the base of the electronics-related categories returns to normal, its growth rate will pick up.
At the same time, supply chain capabilities will also ease the pressure of price hikes to a certain extent.
Xu Ran stated that JD is continuously strengthening its own supply chain capabilities, hoping to more effectively cope with and buffer the price increase of electronics categories through forward-looking layout and relatively flexible inventory adjustment, so that JD can maintain operational resilience while continuously providing users with products with competitive prices.
It is worth noting that Xu Ran also specifically emphasized the product innovation brought by AI.
She said that the rapid development of AI technology has actually brought opportunities for product and category innovation. JD is also working with many brands to respond quickly and develop new products, adopting new technologies rapidly to meet the ever-changing needs of users and even create new demands.
It is worth mentioning that JD's JoyInside has now reached cooperation with nearly 200 brands, with the core focus on providing intelligent interaction capabilities for hardware such as home appliances and robots, thus using JD's AI capabilities to create a smarter and more convenient experience for users.
In terms of general merchandise categories, Xu Ran also expressed confidence in growth.
She said that the core driving force for the growth of general merchandise categories lies on one hand in the self-operated capability and supply chain capability of the categories, including commodity supply, expansion and optimization, price competitiveness building, and service level improvement; on the other hand, it is the continuous growth of users. New businesses such as JD Food Delivery and Jingxi have also brought incremental traffic and new users to JD, creating space for user conversion and cross-purchase.
At the same time, Xu Ran also clearly emphasized that JD will continue to introduce high-quality merchants and cultivate emerging brands, helping merchants achieve refined operations in all links, so that merchants and brands can achieve certain growth on the JD platform while bringing incremental sales.
It is worth noting that judging from the performance of the past three quarters, the sales growth of third-party merchants on the JD platform has continued to rise, and the current growth rate has exceeded that of self-operated categories, while the sales proportion of third-party merchants also increased quarter-on-quarter in the second quarter — this data shows that JD's platform ecosystem is also further optimized.
In addition to the above key businesses, JD also regards the overseas market as an incremental market.
During the earnings call, Xu Ran, CEO of JD, stated that the core competitiveness of JD's Joybuy lies in replicating JD's years of deep-rooted supply chain capabilities overseas and landing them in Europe. In the electronics-related categories, through services such as self-operated retail, logistics fulfillment and on-site installation, Joybuy has gradually won user reputation in Europe, which not only drives the steady increase of user retention rate, but also promotes the business to double its revenue within two quarters.
However, Xu Ran emphasized that Joybuy is still in the early stage of capacity building. Its investment in the second quarter increased slightly quarter-on-quarter, and the loss rate has improved significantly — it is expected that as the business expands and scale effects emerge, the entire financial model will continue to improve.
In addition, for the food delivery business that the outside world is more concerned about, JD stated that the order volume of the food delivery business in Q2 maintained healthy growth, with year-on-year loss reduction of more than 50%; at present, the per-order profit and loss UE of JD's food delivery business has been significantly optimized, and scale growth has also brought about improved delivery efficiency.
In the long run, JD stated that it will accelerate the in-depth collaboration between the food delivery business and core businesses, continuously tap more collaborative value of the food delivery business in JD's ecosystem, and drive the healthy growth of JD's overall user base and revenue as well as efficiency improvement.
When AI becomes a "top leader's priority project"
Looking at the overall development of JD Group, a clue that deserves increasing attention is: The importance of AI in JD's business system has been emphasized to an unprecedented degree.
A typical example of this is: in the second quarter, JD's R&D expenditure reached 7.3 billion yuan, up 37.7% year-on-year, and the expense ratio rose from 1.5% to 2.1%. That is to say, in the same quarter when marketing expenses shrank by 24.8%, JD's R&D investment rose by nearly 40%.
Moreover, judging from the overall data of the first half of this year, JD's total R&D investment has increased by 53.2% year-on-year.
Regarding this investment, Shan Su, CFO of JD, clearly stated during the earnings call that R&D investment is focused on "scalable AI applications", which is regarded as a long-term means to "optimize the overall expense structure".
Of course, another key point for JD to shift its strategic focus to AI is Richard Liu's personal emphasis and increased involvement in the role.
In 2025, the JD Exploration Research Institute was upgraded from a subordinate department of JD Technology to a group directly affiliated institution, and Richard Liu, Chairman of JD Group, personally served as the dean of the JD Exploration Research Institute, which fully demonstrates JD's high attention to AI.
From the current overall business layout perspective, JD's AI has been implemented in multiple scenarios, covering not only digital AI but also physical AI.
For example, in terms of models, JD's model matrix has been further improved. It open-sourced the long audio and video generation framework JoyAI-Echo, the real-time video vision-language interaction model JoyAI-VL-Interaction, and the real-time streaming video editing model JoyAI-Video-Edit. These models are deeply rooted in the industry and released to the industry in an open-source manner for more developers and industry partners to use.
In terms of physical AI, JD is building the world's largest embodied data collection center, and the country's first embodied data collection community has been launched in Suqian. It has now open-sourced EgoLive, the industry's largest human-perspective dataset. The data from real business scenarios such as retail, logistics, healthcare and industry is designed to promote the development of the embodied intelligent industrial ecosystem.
In addition, JD's RoboBase project started construction in Guangzhou in the second quarter, relying on JD's business ecosystem to build industrial infrastructure for the full life cycle of robots. It plans to lay out more than 80 RoboBase robot bases across the country in the next 5 years.
On the logistics side, JD is also leveraging its own advantages to actively deploy AI robots.
As of the second quarter, JD's self-developed "Smart Wolf" goods-to-person solution has been deployed in more than 60 warehouses around the world and landed in the UK and Germany; the "Heterogeneous Wolf" robotic arm can complete identification, grasping and palletizing in 10 seconds; thousands of unmanned delivery vehicles operate on a regular basis in more than 20 provinces across the country; the drone village access network covers 78 administrative villages, with the fastest mountain delivery time of 7 minutes. The system that drives real-time decision-making of this system is JD Logistics' "Super Brain" large model.
In terms of service industry, JD's digital human JoyAvatar has served more than 80,000 merchants in total. The number of digital human live-streaming accounts in Q2 increased by 3 times year-on-year, and the average daily active live-streaming accounts surged by 6 times. During the 618 shopping festival, both the live-streaming duration and transaction value of digital humans increased by more than 4 times year-on-year.
On the retail side, JD also launched the industry's first AI hardware talent show "Aidol Creation Camp" with its partners. More than 50 new products were unveiled through the Creation Camp and made their debut on JD, with the transaction volume in the second quarter increasing by more than 70% quarter-on-quarter.
Overall, JD's AI layout is significantly different from the layouts of other Internet giants in the AI field.
Specifically, JD does not pursue competition with other players on the parameter indicators of basic models. It usually invests based on its own business needs, and pays more attention to business demand-driven and scenario implementation. Its core logic, in JD's own words, is: to build a world-class JoyAI basic model matrix, promote AI to move from the digital world to the physical world, and build a decentralized AI entry and embodied general intelligence.
Of course, behind this logic, there is also a potential goal: to enable AI to better serve JD's business growth.
In any case, after a series of business layouts, in the process of continuously seeking new growth space, the importance of AI has once again received unprecedented emphasis; and JD's strategic shift towards AI is becoming clearer — but the implementation of AI is not easy, and using AI to drive business growth and financial improvement is destined to be a long-term process, and JD still needs more time.
This article is from the WeChat public account "Timelines", author: Wang Zhi, published with authorization from 36Kr.